[Bug c++/105221] New: gcc rejects true ? [](auto) noexcept {} : [](int) {}

[Bug c++/105221] New: gcc rejects true ? [](auto) noexcept {} : [](int) {}

[hewillk at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

            Bug ID: 105221
           Summary: gcc rejects true ? [](auto) noexcept {} : [](int) {}
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: hewillk at gmail dot com
  Target Milestone: ---

int main() {
  true ? [](auto) noexcept {} : [](int) {};
}

GCC-trunk only accepts the above code in C++14 but rejects it in C++17/20.
https://godbolt.org/z/7r4dKbo73

But it seems like it should be well-formed, you can see the explanation in SO:
https://stackoverflow.com/questions/71826315/ternary-operator-applied-to-different-lambdas-produces-inconsistent-results

[Bug c++/105221] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14)

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |c++-lambda
   Last reconfirmed|                            |2022-11-08
            Summary|gcc rejects true ? [](auto) |gcc rejects true ? [](auto)
                   |noexcept {} : [](int) {}    |noexcept {} : [](int) {} in
                   |                            |C++17+ (works for C++14)
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed.

[Bug c++/105221] [10/11/12/13 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14)

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|gcc rejects true ? [](auto) |[10/11/12/13 Regression]
                   |noexcept {} : [](int) {} in |gcc rejects true ? [](auto)
                   |C++17+ (works for C++14)    |noexcept {} : [](int) {} in
                   |                            |C++17+ (works for C++14)
      Known to work|                            |6.1.0
           Keywords|                            |needs-bisection
      Known to fail|                            |7.1.0, 8.1.0

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Worked in GCC 6 even with an explict -std=c++17 .

[Bug c++/105221] [10/11/12/13 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14)

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mpolacek at gcc dot gnu.org

--- Comment #3 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Started with r241944 most likely.

[Bug c++/105221] [10/11/12/13 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14)

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |10.5

[Bug c++/105221] [10/11/12/13 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14)

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org,
                   |                            |jason at gcc dot gnu.org,
                   |                            |ppalka at gcc dot gnu.org

--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
P0012R1 added what is now http://eel.is/c++draft/temp.deduct.conv#5.2 , at that
point:
``If the original A is a function pointer type, A can be "pointer to function"
even if the deduced A is "pointer to noexcept function".
If the original A is a pointer to member function type, A can be "pointer to
member of type function" even if the deduced A is "pointer to member of type
noexcept function".''
but I don't see in pt.cc anything keyed on DEDUCE_CONV that would deal with
this.

[Bug c++/105221] [10/11/12/13 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14)

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P2

[Bug c++/105221] [10/11/12/13 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14) since r7-4383-g51dc660315ef83dc

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |marxin at gcc dot gnu.org
            Summary|[10/11/12/13 Regression]    |[10/11/12/13 Regression]
                   |gcc rejects true ? [](auto) |gcc rejects true ? [](auto)
                   |noexcept {} : [](int) {} in |noexcept {} : [](int) {} in
                   |C++17+ (works for C++14)    |C++17+ (works for C++14)
                   |                            |since
                   |                            |r7-4383-g51dc660315ef83dc
           Keywords|needs-bisection             |

--- Comment #5 from Martin Liška <marxin at gcc dot gnu.org> ---
(In reply to Marek Polacek from comment #3)
> Started with r241944 most likely.

Yes, it started with r7-4383-g51dc660315ef83dc.

[Bug c++/105221] [10/11/12/13 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14) since r7-4383-g51dc660315ef83dc

[jason at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

--- Comment #6 from Jason Merrill <jason at gcc dot gnu.org> ---
Reduced:

void (*p)(int) = [](auto) noexcept {};

[Bug c++/105221] [10/11/12/13 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14) since r7-4383-g51dc660315ef83dc

[jason at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |jason at gcc dot gnu.org
             Status|NEW                         |ASSIGNED

[Bug c++/105221] [10/11/12/13 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14) since r7-4383-g51dc660315ef83dc

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

--- Comment #7 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The trunk branch has been updated by Jason Merrill <jason@gcc.gnu.org>:

https://gcc.gnu.org/g:85131af0603c0af2aa6b40de6cc929905f22bd50

commit r13-6948-g85131af0603c0af2aa6b40de6cc929905f22bd50
Author: Jason Merrill <jason@redhat.com>
Date:   Thu Mar 30 11:07:17 2023 -0400

    c++: generic lambda and function ptr conv [PR105221]

    We weren't properly considering the function pointer conversions in
    deduction between FUNCTION_TYPE; we just hardcoded the
    UNIFY_ALLOW_MORE_CV_QUAL semantics, which are backwards when deducing for a
    template conversion function like the one in a generic lambda.  And when I
    started checking the ALLOW flags, I needed to make sure they stay set to
    avoid breaking trailing13.C.

            PR c++/105221

    gcc/cp/ChangeLog:

            * pt.cc (unify) [FUNCTION_TYPE]: Handle function pointer
            conversions.

    gcc/testsuite/ChangeLog:

            * g++.dg/cpp1z/noexcept-type27.C: New test.

[Bug c++/105221] [11/12/13/14 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14) since r7-4383-g51dc660315ef83dc

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|10.5                        |11.5

--- Comment #8 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC 10 branch is being closed.

[Bug c++/105221] [11/12 Regression] gcc rejects true ? [](auto) noexcept {} : [](int) {} in C++17+ (works for C++14) since r7-4383-g51dc660315ef83dc

[jason at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105221

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
   Target Milestone|11.5                        |13.0
         Resolution|---                         |FIXED

--- Comment #9 from Jason Merrill <jason at gcc dot gnu.org> ---
Fixed in 13.  The fix seems a bit risky for backporting, but I'm open to it if
there's any interest.